• English
  • Deutsch
  • Log In
    or
  • Research Outputs
  • Projects
  • Researchers
  • Institutes
  • Statistics
Repository logo
Fraunhofer-Gesellschaft
  1. Home
  2. Fraunhofer-Gesellschaft
  3. Konferenzschrift
  4. Batch-wise Regularization of Deep Neural Networks for Interpretability
 
  • Details
  • Full
Options
2020
Conference Paper
Titel

Batch-wise Regularization of Deep Neural Networks for Interpretability

Abstract
Fast progress in the field of Machine Learning and Deep Learning strongly influences the research in many application domains like autonomous driving or health care. In this paper, we propose a batch-wise regularization technique to enhance the interpretability for deep neural networks (NN) by means of a global surrogate rule list. For this purpose, we introduce a novel regularization approach that yields a differentiable penalty term. Compared to other regularization approaches, our approach avoids repeated creating of surrogate models during training of the NN. The experiments show that the proposed approach has a high fidelity to the main model and also results in interpretable and more accurate models compared to some of the baselines.
Author(s)
Burkart, Nadia
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB
Faller, Philipp M.
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB
Peinsipp, Elisabeth
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB
Huber, Marco
Fraunhofer-Institut für Produktionstechnik und Automatisierung IPA
Hauptwerk
IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, MFI 2020
Konferenz
International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI) 2020
Thumbnail Image
DOI
10.1109/MFI49285.2020.9235209
Language
English
google-scholar
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB
Fraunhofer-Institut für Produktionstechnik und Automatisierung IPA
Tags
  • Explainable Artificia...

  • Künstliche Intelligen...

  • maschinelles Lernen

  • neuronales Netz

  • Cookie settings
  • Imprint
  • Privacy policy
  • Api
  • Send Feedback
© 2022